Tom Watson, the founder of IBM, is quoted as saying, “If you don’t genuinely like your customers, chances are they won’t buy.”

In any technology business, particularly businesses servicing non-technical customers, it is tempting to assume “user error” anytime a customer calls your support line. Most Software as a Service (SaaS) businesses would fall into this category of thinking. If your service team begins each day assuming the customer is generally wrong, naive, or just too lazy to read the manual, then they will treat your customers as the enemy. This is not acceptable.

Great businesses treat every customer as a friend or colleague. Great businesses view a customer with a problem as an opportunity–an opportunity to learn how to make the product or service better. It is an opportunity to up-sell additional services. It is an opportunity to build a raving fan.

A raving fan is much more than a customer. Let me explain with an example. Imagine you are out for dinner. The server comes to your table and asks, “How was your meal?” If you simply reply, “Fine,” then you are not a raving fan. But if instead you say “Great” and then, as you are pulling into your driveway later that night, you bump into your neighbor and shout out to him, “John, I just ate at a great restaurant,” then you are a raving fan.

To promote true customer partnerships I recommend beginning every customer support call with the assumption that your product or service could be better and that the customer knows what they are talking about, because if your product was perfect the customer would not need to call. If, in every interaction, you treat your customer as an intelligent thoughtful person, as a friend, and always assume a customer problem is an opportunity to improve your product and processes–you will build raving fans. Raving fans will promote your business.

Make a customer a friend and a raving fan today. Grow your business.

Bookmark this post:

…continued from NBA Playoff Laptop Flap…

While not all coaches and front offices are embracing sports analytics, a few major franchises are, including the Boston Celtics, Oklahoma City Thunder and San Antonio Spurs of the NBA (all still-alive playoff teams, by the way).

Sports analytics are big enough now that there’s an annual conference—the MIT Sloan Sports Analytics Conference. And that conference actually draws some big names in sports.

The speaker list for 2012’s conference (held in March) included reps from various franchises, sports and the sports media. Some names…Dallas Mavericks owner Mark Cuban, ESPN commentator Michael Wilbon, NASCAR IT chief Andy Schwalb, ESPN columnist Bill Simmons, Manchester United CEO David Gill and NHL Commissioner Gary Bettman, just to name a few.

So, what exactly is sports analytics? According to Kevin Goodfellow of Sports Data Hub, it’s the science of collecting and culling usable information from mass amounts of data, or “connecting fragmented data into a form that enables deeper analysis.”

To Ben Alamar, NFL consultant and Director of Analytics for the OKC Thunder, it’s seeing the forest versus just the trees. For example, every week assistant NFL coaches gather the statistical data they deem important and present it to the head coach, who makes strategy based on the data.

The problem, according to Alamar, is that the assistants only focus on their own areas (i.e., special teams, defensive secondary, offensive line) without seeing the whole.  And they may miss connections to what the other assistants are working on. It’s about managing the team’s data as a whole “to gain important insight from the combination of all” of the parts.

Sports analytics can be high-level or low-level, from something like time of possession in football to which hand a point guard in basketball dribbles better with (and by what degree that hand is better). It’s up to the franchise to decide how far they want to go with analytics.

A simple example is the one offered by ESPN analyst John Hollinger (creator of the Player Efficiency Rating system, which statistically rates NBA players on per-minute productivity). During their 2011 championship playoff run, the Dallas Mavericks changed their lineup because the stats said they were better with J.J. Barea in the lineup. They played him more and eventually won the NBA Finals.

More complicated is former NBA coach and current ESPN commentator Jeff Van Gundy’s example of asking a guy to shoot a three-pointer when he’s 0 for 3. Statistically, he may be likely to make the next one, but if he misses again it might affect his confidence for the rest of the game.

To Van Gundy, sports analytics helped him better understand his team as well as test and verify gut feelings, although he still had to temper his use of the information.

And therein lies the rub. Some people think the “bean counters” instead of the coaches will be making in-game decisions if analytics continues to grow. Others who know coaches and the pressures they face don’t believe coaches will gladly turn over control. We’ll see, I guess.

Bookmark this post:

In Game 5 of the NBA playoff series between the Denver Nuggets and Los Angeles Lakers last week, the Nuggets had a laptop with them on the bench in the closing moments of the game.

The Lakers complained after the game that NBA rules forbid the use of laptops or tablets during play. That’s not actually true, but the flap brought attention to a growing trend in professional sports—the use of technology.

The movie Moneyball dramatizes the real-life use of analytics by Oakland A’s GM Billy Beane. Faced with a small budget and the loss of key players to free agency, Beane turns to statistical analysis to field a team—and win. The movie is basically the general public’s introduction to a movement that has been gaining strength in the last few years.

Sports analytics aren’t brand new, but while they’ve yet to take a firm hold on pro sports teams, they’re making inroads. Sports analytics is basically using mass amounts of data to get a competitive advantage.

In professional cycling, for example, analytics can take the form of cranksets and wheel hubs that measure a rider’s power output (in watts) during training and races. In basketball, it can take the form of laptops on the bench during games.

Mike Bresnahan of the Los Angeles Times after last week’s LaptopGate—

The Lakers were privately seething after seeing the Nuggets use a laptop computer in their huddle during a 20-second timeout with 19.9 seconds left to play.

The computer apparently belonged to an assistant coach sitting behind the bench with it. NBA rules forbid the use of such devices in the huddle, which won’t change the final score but can carry a hefty fine of up to $250,000.

Actually, the Nuggets didn’t use the laptop but would have been within the NBA’s rules to do so. Nuggets head coach George Karl said the team kept the Lakers’ “end-of-game plays” on a laptop so they could review them come crunch time.

“It is permissible to use a laptop or tablet (i.e., iPad) for the purposes of accessing, using or presenting statistical and scouting information to players and coaches during games,” NBA spokesperson Tim Frank told Yahoo! “This can include video or photographs of prior games.”

Basically, teams can use laptops or tablets as long as they’re not looking at stuff from the game being played. But that’s just the tip of the iceberg for sports analytics, which may transform professional sports.

Stay tuned for the rest of this post, Sports Analytics Emerge…

Bookmark this post:

The office hours are now over, but you can watch the replay on Spreecast!

Join our live developer office hours on Spreecast right now where you can grab instant answers and direct feedback on your Twilio questions.

We invite you, our noble community of Twilio hackers, to ask questions, talk about your favorite hacks, tell us what features you’d love to see, or just talk about Star Wars (disclaimer: no Jar-Jar).

Twilio Office Hours Streaming Live Now

Bookmark this post:

The reason people typically prefer interactive voice response, or IVR, systems that are equipped with features like automatic speech recognition and text-to-speech is because they allow for more natural, organic interactions between the person and the application.  Adding personalization and humanization to IVR systems often times offers the end user a more personal, satisfying experience.

Instead of requiring that users submit data and information using DTMF or numerical entry, speech recognition allows users to speak their answers like they would if they were speaking to a human operator.  The interaction is less clinical and uniform, and end users typically find that enjoy a more enjoyable experience.

Almost universally, companies try to offer customers a personalized experience that engenders brand loyalty and positivity towards their product.  Google is taking note of this trend, and has recently announced a new service called Knowledge Graph.

Knowledge Graph will significantly alter how search results are displayed, with Google claiming that this new feature will “make their search engine think more like a human”.

So how does it work?  According to a CNN article, Knowledge Graph results will be arranged according to categories that the search term has been associated with.  The results will appear in boxes, and users can further click on these boxes to get results about the precise topic they are seeking information about.

Google will categorize all the existing knowledge on a specific search term that they possess into visually representative images.  The user can quickly and easily look over these categories, choose the search results that most accurately represent their desired search term, and receive highly accurate data based on their specified click through information.

In addition, the Knowledge Graph will take in to account auxiliary information like recent searches, location, and preferences to generate search terms that are topically relevant to each individual person.

On the one hand, the idea is novel, and will inevitably change the way Internet searches are performed for every company if the concept is well received and successful. On the other hand, it means that Google will be storing personal information on each and every user, including (but probably not limited to) location, previous search terms, and previous websites visited.

While it may be convenient to receive personally tailored search results almost instantaneously at the click of the button, the personal data and privacy sacrifices that this type of search enables might be very steep.  Just how long will user data be stored?  Additionally, in what manner is Google intending to store this type of data?  Cue the data security and privacy advocates, it sounds like we might have a bit of a fight on our hands…

Bookmark this post:

This post was originally posted on the Startup Weekend blog, found here. Carter Rabasa is a Twilio Developer Evangelist based in Seattle and he’s looking forward to meeting all the local developers so say hello at carter@twilio.com.

Twilio is gearing up for a big weekend in Seattle as we’re sponsoring Seattle Startup Weekend and I’m excited to spend this time with the local tech community. Twilio was started in Seattle, so this city figures deeply into our DNA and since I’m based here, my objective was clear:  find the startups in Seattle, who is supporting them and how I can help. I’m thrilled to announce that Twilio will be hosting regular office hours in The Easy, located in the basement of Founder’s Co-op, every Wednesday, from noon-2pm.

The purpose of these office hours are for startups to get their technical and business questions answered.  Feel free to bring your code, show off your prototypes and work through your ideas while meeting other Twilio developers in Seattle.  You can just show up or sign-up for a slot.

Despite having lived in Seattle for over two years, I was just beginning to uncover the local tech community. After joining the Twilio team I spent the first few months getting to know the different places that entrepreneurs and developers spend time.  These ranged from co-working spaces to incubators to coffee shops.  I also got to know many of the people who support and provide services to startups.  I was impressed by what I saw, but also realized that Seattle has a ways to go to provide the same kind of ecosystem that startups enjoy in places like San Francisco and New York.  Still, one place that I visited really blew my socks off.

In March I had the chance to meet Chris Devore.  Chris and his business partner Andy Sack operate Founder’s Co-op.  FC is both a seed-stage fund and  one of the most vibrant spaces for startups and entrepreneurs in Seattle.  In addition to housing FC-funded companies, FC plays host to TechStars Seattle from August to November. After talking to Chris, it was obvious that FC was the logical home for the Twilio Seattle office.  So, in April we moved in!

I am really excited about Twilio’s expanding presence in Seattle and our new home in Founder’s Co-op.  Every time I turn around, a new building has gone-up in South Lake Union or a new incubator has popped-up downtown.  More and more people are taking the plunge and starting companies, which means that companies that support this (like Twilio and Startup Weekend) have a lot more people to help.  We’re looking forward to it!

Twilio Makes a Comeback in Seattle with Office Hours & Startup Weekend

Bookmark this post:

Automatic speech recognition (ASR) technology is an integral feature of many IVR systems.  ASR enables a user to speak their answers as opposed to numerically entering them, permitting for much greater depth and quality of answers offered.

A study conducted by researchers from the Hammill Institute on Disabilities utilized ASR technology to track babies’ early language development by recording, transcribing and analyzing the amount and type of language queues they were exposed to within their home environments.

The findings were in keeping with previous studies, showing that children of the most talkative parents were the ones producing more vocalizations and conversational turns.  This learning and speech pattern continued and, in fact, increased month over month.

The data reflects the idea that children’s exposure to language at home significantly affects the rate at which they learn to talk.  Analyzing the data, researchers also found that there is no “normal” when it comes to home learning environments, and that each child’s situation varied in terms of their exposure to language.

Other interesting findings?  Adult females provided the majority of the words heard by the child per day, but this could be due in part to the fact that the mothers included in the study reported spending significant time at home with their baby, more than most average mothers.

The technology utilized in this study was quite unique.  The ASR was able to accurately distinguish between adult and child vocalizations.  The fact that the speech recognition engine could easily and seamlessly separate adults and children’s voices offers a variety of possibilities for those collecting and processing data based on vocalizations.

Referring back to my write-up from last week, if speech recognition technology is able to separate adult and children’s voices, it could also feasibly measure less quantifiable aspects of the human voice that are equally important in this type of data collection, specifically emotion and intonation.

The bottom line is that ASR technology enables research advancement through its use.  Research on “autism, effects of television, language learning in linguistically diverse environments and language intervention” are all possibilities that become realistic through the use of this type of system.

Bookmark this post:

Martin Pannier was looking for a way to bring the power of search similar to Siri to a wider audience, why should iPhone 4s owners have all the fun? What he built is WolfraSMS: a simple SMS app that harnesses the search capabilities of Wolfram|Alpha, powered by Twilio.

Martin is the co-founder and COO of SocialFolders, a service to organize your social content in the cloud. WolfraSMS is the very first app Martin built with Twilio – quite impressive. Follow Martin on Twitter @martinpannier and read his blog here.

What exactly is WolfraSMS?

WolfraSMS is—very simply—a way to ask the Wolfram|Alpha computational search engine questions, by sending a simple SMS. You can send a variety of questions, such as “Where was Obama born?” or “Who was the third president of the United States?” to the WolfraSMS number and it will send you back the answer in SMS form.

What was the inspiration for building WolframSMS?

I love Siri. Being able to ask questions and get the answer, and not only a series of links, is quite remarkable. But Siri is only available to iPhone 4S users. I wanted to bring some of the power of Siri to the 99%!

SMS is incredible for its ubiquity: anywhere, and with any phone, you can send and receive SMS messages. Geeks don’t quite realize the power of this ubiquity, as they always tend to have huge data plans and sophisticated smartphones.

Match the two… boom: WolfraSMS.

How are you using Twilio?

Twilio powers the whole SMS hub.

What other technologies are you using to build/support it?

The app was coded in Python and is hosted on Heroku.

How did you get started developing with Twilio?

With WolfraSMS :) It’s actually my first Twilio app.

What’s next for WolframSMS?

This is a side project for me, so with limited time on my hands, what’s realistically next is easter eggs and more thorough use of the Wolfram|Alpha answers—as for brevity reasons, only the most relevant part of the answers is returned. If I had more time, I would include the same search engines as Apple uses for Siri—starting with Yelp.

Then Siri would really be available to anyone in the world!

WolfraSMS, the Power of Search via SMS and WolframAlpha

Bookmark this post:

Eight firms settle patent lawsuits with Klausner, three more suits remain.

Bookmark this post:

Educational gaming app connects English and Spanish learners

Bookmark this post:

Automatic speech recognition (ASR) is one of the critical elements of IVR applications.  ASR technology enables an interactive voice response, or IVR, system to process user input through speech, as opposed to traditional methods of data submission like DTMF.

With speech recognition, users can clearly and succinctly communicate with an intelligent application, and have the application respond in turn.  The technology is novel and has required a significant degree of fine-tuning since its initial introduction, but when deployed properly, the application’s ability to recognize, process and ultimately understand spoken language is very impressive.

The technology has been deployed in a variety of systems, interacting with a wide-ranging demographic of end users.  Some of the most interesting uses of ASR are in behavioral tracking.  Individuals attempting to understand or solve a particular problem interact with ASR applications in a natural, unassuming manner, giving researchers valuable insight into their thoughts, feelings and psyche.  This data research is typically highly accurate and very honest because users are interacting with a machine instead of a person and often don’t feel compelled to modify their behavior.

An article from the Communication Disorders Quarterly dissects how the dialogue and vocabulary children are exposed to in their home environments affect their verbal skills and comprehension from a very young age.

The researchers behind the article conducted a study over a period of ten months of children between twelve and twenty months old.  They collected audio recordings in the children’s homes in twelve-hour increments (which greatly expanded on previously conducted research from the 1970s and 1980s), and did so on a weekly basis.

Small digital recorders placed in specially designed clothing pockets enabled subjects to easily transport the device throughout their daily activity.  The recordings were analyzed by automatic speech recognition technology and transcribed.  Pre-selected variables, including adult word count, child vocalizations and conversational turns were input into a database for further analysis.

There were significant variations in the amount and type of talking done in each of the homes studied. The research took into account the amount and variety of words addressed to and heard by a child each day.  There were also wide variations in the patterns of talk that occurred over months. As the child grew, there were a larger number of responses and words spoken to them and by them.

Patterns of talk were also varied, with the most talking being done (by both adults and children) “in the early morning starting at 7, and in the late afternoon-early evening.  The pattern was lowest during the noon and midday hours.  Talk dropped to near zero at 7pm.”

So, did the quality and quantity of spoken words significantly affect the children being studied?  And if so, how?

Stay tuned for Babies Hear Everything…

Bookmark this post:

Interactive voice response, or IVR, applications are typically commissioned and deployed with the intent of positively solving an existing issue or facilitating further communication abilities via technology.  Most, if not all IVR systems are developed with the hopes of enabling additional and alternative communication channels for end users seeking to interact with technology in a constructive manner.

While IVR applications are productively integrated into business environments, incorporating backend databases and streamlining incoming and outgoing data collection processes, there are other more personalized uses for the technology as well.

Specifically in the field of healthcare, there’s a host of ways interactive voice response technology can benefit individual users.  From smoking cessation programs where users place calls to both give and receive data and information regarding their progress and challenges, to subjects who suffer from chronic pain who want to keep a thorough record of their experience and progress, IVR systems can assist individuals in their personal lives with efforts to monitor and maintain their personal health.

A study published in the Communication Disorders Quarterly set out to assess children’s home language environments using ASR (automatic speech recognition) technology.  Researchers Hart and Risley conducted a research study in the 1970s and 80s with the end goal of “improving the language, academic, and social achievements of urban children living in poverty.”

They conducted the research via monthly audio recordings collected from children’s homes.  One hour of audio was recorded and observed each month, starting from when the children were 10 months until they were 36 months.

Their findings show that a child’s home language environment (specifically pertaining to the number of words that are addressed to them by adults) greatly effect how a child learns and retains spoken vocabulary, in addition to other language indicators like vocalizations and conversational turns.

Thirty years later, researchers from the Hammill Institute of Disabilities attempted to expand on the study’s findings in the hopes of understanding exactly how a child’s language environment affect their language development and efficiency.

How did they do this?  Instead of sampling one hour per month, they sampled twelve hours a day, one to three times per week.  They also explored gender differences in the adult vocalization heard by children and compared home language patterns to the commonly recognized measures of child development and language.

So what exactly did they find?  Stay tuned for No More Baby Talk…

Bookmark this post:

Voxeo China is hiring a Senior Java Software Engineer! We’re looking for someone with Java based system design, development, testing, documentation, and maintenance, and the ability to communicate and collaborate with our engineers in China, Germany and the US.

Desired skills include:

• Bachelor/Master Degree in Computer Science/Information Technology;
• At least 3 years of full-time working experience in building Java based products;
• Must be experienced in multi-threading programming, networking and OO design/development;
• Good command of English reading and writing, plus if able to communicate verbally;
• Knowledge about VoIP related technologies is a plus;
• Solid communication skills to clearly articulate technical issues in both written and verbal form;
• Must be strongly self-motivated on learning new technologies;
• Must be a good team player.

Interested? Get more info or apply via LinkedIn.

©2012 Voxeo Talks. All Rights Reserved.

Related posts:

1. We’re hiring – looking for a Java guru…
2. We’re Hiring! Java Developers, NOC Engineer, Web Developer…
3. Are you a Java guru? WE WANT YOU! (Voxeo’s hiring!)

Bookmark this post:

Following the February launch of Twilio Client for iOS, we are very excited to expand our mobile tools with the launch Twilio Client for Android. This makes it possible for any Android application to make and receive voice calls over a cellular data or WiFi connection without ever using standard cell phone minutes.

With this launch, Twilio Client is now available for Android, iOS and web browsers to implement VoIP communications into any mobile app. Seamlessly talk between these platforms and traditional phones to let your users talk over whatever medium they choose.

Building Apps That Talk in the Cloud

Using Twilio Client for Android, developers have access to a few key features that will make building communications apps easy.

• Real-time Presence – Build buddy lists into any app letting users know who is online and set up notifications when a user is available for voice chat
• App Backgrounding – The SDK can receive voice calls even when a different app is in use, letting users switch between apps easily to respond to incoming calls.
• Cross-Platform Interoperability – Twilio Client supports Android, iOS, and web browsers so apps can make calls between any platform, including bridging VoIP calls with traditional phone calls.
• Voice Over Data and WiFi connections – Make and receive calls over data and WiFi connections without using carrier minutes

Getting Started with Twilio Client for Android

Ready to get started on building a mobile app that makes and receives phone calls?  We’ve got documentation, helper libraries, sample code, and an active community of developers ready to answer your questions.

• Get Started with Twilio
• Twilio Client
• Twilio Client for Android Quickstart
• Download Android SDK

If you are looking for iOS information take a look at the Twilio Client iOS Quickstart, iOS User Guide and download the SDK here.

Need more help?  Can’t code but want to find someone who can?  That’s cool!  Contact us at sales@twilio.com and we’ll connect you with members of our developer community who can help. Can’t wait to see what you build.

Twilio Client Expands to Android to Enable Cross-Platform VoIP Calls

Bookmark this post:

Project involves machine translation

Bookmark this post:

Security has long been the singular focus of voice biometrics applications. And for good reason. A voice print, is as unique to an individual as a fingerprint or iris scan. And since speaking is what we do naturally, voice is as user-friendly as it is secure. Voice biometrics implementations are on the rise, Nuance customers alone have surpassed 15 million voiceprint deployments, up nearly 90% from the previous year, which has been driven primarily by government legislation and corporations'...(read more)

Bookmark this post:

…continued from Translation with ASR…

The implications of using speech recognition for translation in the interactive voice response, or IVR, world could be far-reaching. IVR systems and translation could go hand in hand.

Barbara Dragsted, Inge Gorm Hansen and Inger M. Mees from the Copenhagen Business School were curious to see whether an automatic speech recognition (ASR) program (vital to interactive voice response) could improve translation among the translation and interpretation students at their school.

The researchers asked students to translate various documents from Danish to English (the more common of directions, as English is a much more widely used language worldwide than Danish). They published their findings in a paper: “Speaking Your Translation: Students’ First Encounter with Speech Recognition Technology.”

Dragsted, Hansen and Mees recorded the audio output of the students speaking to the ASR and analyzed speed of translation, quality and the types of mistakes made. They found that the time for translation was reduced by the use of an ASR without hurting quality.

They also found that a majority of the misrecognitions of the ASR could be placed on the students. When you think about it, that’s not a real surprise, given the variables confronted by the ASR.

It’s not a surprise to someone who works in IVR, either. There’s a lot of nuance in understanding the spoken word that humans grasp automatically from years of experience.

…we looked into the number and types of error that occurred when using the SR software. Items that were misrecognised (sic) by the program could be divided into three categories: homophones [words that we pronounce the same and may spell the same, although they have different meanings], hesitations and incorrectly pronounced words. Well over fifty percent of the errors were caused by students’ mispronunciations.

I could see that. Think about all the variables between students’ different accents (think of a northern American accent versus a southern American accent), the way they pronounce certain words (even people from the same region may pronounce words differently) and mistakes.

In the end, translation with ASR was somewhere in the middle of spoken sight translation and written translation as far as time spent on the task (written the longest) and quality (written the highest quality). But the researchers feel that ranking could change.

We hypothesise (sic) that with more practice and training, SR time consumption will approach that of sight translation, and SR quality will approach that of written translation.

That interpretation points towards improvements in the translators as opposed to the speech recognition software, which has improved in leaps and bounds in the last few years, as evidenced in the countless IVR systems now using it effectively.

Bookmark this post:

Translation is an area that could emerge in the interactive voice response, or IVR, world soon. Although I’ve actually worked as a translator (written French to written English), I’m surprised I haven’t thought more about it.

Interactive voice response relies heavily on automatic speech recognition (ASR). With translation added to the equation, IVR may rely on it even more. Maybe one day we’ll have IVR systems that automatically understand and converse in 100s of languages.

For now, though, a group of Danish researchers has been looking into using ASR for translation. What they’ve found is that automation could greatly benefit translators as they do their work.

Barbara Dragsted, Inge Gorm Hansen and Inger M. Mees from the Copenhagen Business School conducted a study using Danish translation and interpreting students from the college to test if using an ASR could help.

The researchers had the students produce translations from Danish into English “under different working conditions: written translation, sight translation [translating orally from written text] and sight translation using a speech recognition program, i.e. software which automatically converts spoken output into written text.”

It’s kind of a combination of translation and dictation, which touches on another of the most widely employed uses of ASR other than in IVR. Dictation has served physicians for centuries, and today many physicians are using tablets or computers with speech recognition software on them to handle dictation for them.

During the study, the researchers were looking for three key pieces of information. They wanted to know if there was any difference in the time it took to perform the translation, the quality of the translations and the type of mistakes made.

Stay tuned for the rest of this post, The Future of Translation?…

Bookmark this post:

Lyle Pratt is a DOer by nature, with a business-focused education on paper, he’s been writing software for years. After missing API Hack Day because of the flu, Lyle reached out to share with us the app he built anyways – talk about hack commitment. What he built was an add on to his current start up BetterVoicemail, an app that allows users to simply build their own custom phone system with voicemail, call routing, virtual phones and more.Meet BuildANumber, a brilliantly easy way to set up a customized phone system that anyone can build.

We interviewed Lyle on how he was able to use his entrepreneurial skills and love for code to create BuildANumber and BetterVoicemail. First, check out the demo of BuildANumber below:

Twilio: What was your inspiration for BuildANumber?

Lyle Pratt: BuildANumber started as a way to make the BetterVoicemail.com back-end more flexible, but it turned into a product in and of itself. I was also pretty frustrated with the limited functionality of existing virtual phone system providers. BuildANumber tries not to limit your phone system’s potential in any way.

BetterVoicemail is a product created specifically for Real Estate Agents. Soon, the back-end code behind BuildANumber will power BetterVoicemail.com, which will let us easily offer tons of additional features on a case-by-case basis if a customer needs them.

BuildANumber is also about opening up Twilio’s services to a larger base. Twilio provides a lot of incredible tools that let developers easily create custom phone and SMS systems, but the vast majority of people can not directly benefit from Twilio’s tools because they don’t know how to code or setup and manage servers. OpenVBX helped open Twilio up to a larger base of people, but you still have to know how to manage and set up a server to use it, or pay someone to set it up for you. With BuildANumber, Twilio’s services are suddenly available to anyone. No programming or technical knowledge required!

T: What technologies did you use to build it and how are you using Twilio?

LP: I built BuildANumber on a Python/Django back-end running on load-balanced EC2 servers connecting to an Amazon RDS Multi-AZ database system. The call template builder uses custom Javascript that interfaces with the back-end and Twilio Client to let customers test their templates on the fly. I use Git for version control and Sublime Text 2 as my editor of choice.

I created BuildANumber to be able to use every Twilio technology available, including TwiML, the REST API, Twilio Connect, and Twilio Client.

T: Aside from the features listed above, what do you want to do next with BuildANumber?

LP: There is still a lot of functionality I want to integrate into BuildANumber. One of the functions I’m most excited about is the addition of SMS templates. Currently BuildANumber only lets you create templates that respond to phone calls. Soon, you’ll be able to create templates that respond to text messages in a similar way.

Another aspect I’m planning to develop is functionality that will let your templates respond to calls differently based on who is calling. As an example, if someone calls who is in your “Family” group, you might want them to have a different experience than someone calling about from your “Business” group.

Android and iOS mobile apps are also in the works that will let you easily access your missed calls and voicemails on the go.

Lastly, I want to increase the level of organization support offered by the product. I want BuildANumber to be friendly to large organizations who might want to be able to do things like share templates across multiple accounts or easily route calls to various organization members.

Lyle has been able to publish a live ALPHA version of BuildANumber using Twilio Connect, and just recently launched a feature that allows you to set up specific call functionalities that respond to each caller differently depending on their number. To hear a simple example created with BuildANumber, call (512) 843-8606.

DOer Lyle Pratt Launches BuildANumber alongside BetterVoicemail

Bookmark this post:

Oaisys AllWays and Gateway to the Cloud will enable users to take advantage of advanced recording anywhere, anytime, and on any device.

Bookmark this post:

Offering brings Web-based speech analytics for the SMB market

Bookmark this post:

Unit acquired by private equity fund

Bookmark this post:

…continued from Wake-Up-Words…

Automatic speech recognition (ASR) is at the heart of all interactive voice response, or IVR, systems. Over the years, speech recognition has made leaps and bounds, and it could be about to make another one.

Florida State University researchers V.Z. Kepuska and T.B. Klein have been working on a way to make it easier for ASR systems to recognize when someone is speaking to them.

Wake-up-words (WUWs) are exactly what they sound like—words to wake the computer up, or command words. Once you wake the computers up, via a command word, interaction can begin.

When describing what a WUW is, the researchers talk about a system that keeps running and only engages when it hears WUWs. The difficult task in that, however, is distinguishing the context of the words.

In their paper, A Novel Wake-Up-Word Speech Recognition System, Wake-Up-Word Recognition Task, Technology and Evaluation, Kepuska and Klein use an analogy of a speech recognition-operated camera system on a movie set.

The system has several command words it reacts to, including lights, camera and cut. When the director says lights, the lights go on; camera, the cameras go on; cut, the scene stops.

However, if one of the lines of dialogue includes the word cut, which is one of the command words, it’s crucial that the system reacts to it only when it’s meant to cut the scene.

Therefore, if an actor says, “I cut the rope,” the system won’t cut the cameras off. However, if the director says, “cut,” the system cuts the cameras. It’s a fine distinction, and one that ASR systems will need to come to grips with if they want to be HAL one day.

Bookmark this post:

Automated speech recognition (ASR) is vital to technologies like interactive voice response, or IVR, systems that handle human-computer interaction via audio. Yet we haven’t quite gotten to the point where they’re like HAL 9000 from 2001: A Space Odyssey.

One of the challenges for ASR systems (and, subsequently, IVR systems) is distinguishing between words spoken to the ASR and words spoken to someone else. Researchers from the Florida Institute of Technology have looked into it and maybe have the next breakthrough.

Even for us (humans, that is), it can be hard to know if someone is talking to us, someone else or even themselves. We use linguistic clues as well as context and other variables to decide whom a speaker is addressing.

If we have trouble with it, how can we expect a speech recognition system to figure it out? The answer may exist with a new idea in speech recognition: wake-up-words.

V.Z. Kepuska and T.B. Klein of the Florida Institute of Technology are hoping wake-up-words (WUWs) combined with increased accuracy will take ASR to the next level.

A WUW is basically a command word that the ASR system will recognize, given the right context. A WUW is just like it sounds—a word that wakes the ASR system up when a user is addressing it.

According to the researchers, a “WUW SR is defined as detection of a single word or phrase when spoken in the alerting context of requesting attention, while rejecting all other words, phrases, sounds, noises and other acoustic events and the same word or phrase spoken in non-alerting context with virtually 100% accuracy.”

Stay tuned for the rest of this post, HAL, Wake Up…

Bookmark this post:

Space trips, rap stars and huge stacks of money.

These might read like the answers to “What tops every 10 year-old’s wish list?” but they’re actually units used to quantify modern-day marketing automation spend. In a new infographic on the evolution of marketing automation (via MarketingProfs), we share what our research had to say on the state of marketing automation and B2B growth.

Marketing automation has moved way past its email origins and now incorporates everything from social media to voice communication. With every means of B2B communication under one big tent through marketing automation, companies who use the technology process have leg up on the competition.

To learn how those space trips and stacks of money make sense in marketing automation, check out the infographic below:

Read the full post here on MarketingProfs:

What the Evolution of Marketing Automation Looks Like [Infographic]

Bookmark this post:

We love Chicago and we don’t get to give the city as much attention as we’d like. But every so often, an entire series of events line up to let us visit the Chicago area to enjoy the only city with real deep dish pizza, connect with developers, and generally get to know the community better.

Jon is on the left.

Starting tomorrow, May 15th, Twilio’s Jon Gottfried will be visiting Code Academy and the Chicago Node.js Meetup. Code Academy is a beginner-focused programming school that teaches students to build web applications with Ruby on Rails. He is running a workshop for their current classes on how to add voice and SMS support to a Rails application and how to build an awesome game of Telephone with Twilio and Rails.. Though the current classes are full, you can still apply to join their next Web Design, Development, and User Experience cohort kicking off in June.

At the Chicago Node.js Meetup that evening, Jon will demo both his Two-factor Authentication Twilio/Node.js application and a Twilio-powered Pacman clone. There is plenty of room so RSVP on their sign up page to join.

Devin is on the right

Next, Devin Rader is hitting Chicago Code Camp on Saturday May 19th. In addition to his session to dissect what goes into designing good REST APIs using the new ASP.NET Web API Stack, you can learn about mobile development with Qt, Test Driven Development with iOS, an overview of Windows 8, and finally Microsoft’s Robotics Studio. Overall, a “full day” is an understatement.

Keith is in the middle

To round out the week you can find me (Keith Casey) on May 22nd-25th at php|tek. This event is near to my heart as I help co-organize and this time I’ll be serving as the facilitator of the Hackathon and Unconference sponsored by our friends at Engine Yard/Orchestra.io and Mashery respectively. Last year, the Unconference ran until midnight with sessions on regular expressions and random hackery while the Hackathon lasted until 2am. Unlike many hackathons, the theme isn’t just a clever hack but draws project leaders from the open source community who build apps they use on a daily basis.  Before you go, catch my closing keynote Thursday morning.

If you can join us for one, any, or even all of these events, please do. I promise t-shirts, stickers, and new ideas that will make you think.

This Week Twilio takes Chicago with Text Messages and Pizza

Bookmark this post:

We are kicking off a series of developer office hours on the Twilio Spreecast Channel this coming Thursday, May 17th at 4pm PST/7pm EST where you can grab instant answers and direct feedback on your Twilio questions.

We invite you, our noble community of Twilio hackers, to ask questions, talk about your favorite hacks, tell us what features you’d love to see, or just talk about Star Wars (disclaimer: no Jar-Jar).

Join me on Spreecast for our first online developer office hours – Thursday, May 17th at 4pm PST/7pm EST to 5pm PST. You can RSVP at http://www.spreecast.com/events/twiliohelp517 or just hop online on the 17th!

A hat tip to Dwolla Evangelist Michael Schonfeld who kicked off office hours the other week on Google+. We thought it was such a great idea we wanted to offer it to our community as well. Maybe one day we’ll do an API mashup developer hangout, what do you think?

Look forward to seeing you there!

Twilio Office Hours: Hang Out in the Cloud With Real Live Twilions!

Bookmark this post:

Dragon Apps expand availability and language support in eastern Europe

Bookmark this post:

Contract to provide enhanced security for mobile devices

Bookmark this post:

…continued from Electric Car Fire…

After the fire became public, Fisker issued a statement, which included some eye-raising comments—

There are conflicting reports surrounding this particular incident. The cause of the fire is not yet known and is being investigated.

I’m not sure if they released that after Baker’s assertion that the Fisker was to blame or not.

Based on initial observations and inspections, the Karma’s lithium ion battery pack was not being charged at the time and is still intact and does not appear to be a contributing factor in this incident.

Yet, last month five people were injured when a prototype battery caused an explosion at a GM battery-testing lab. The battery was made by A123, the same company that provides batteries for the Fisker Karma. According to the Detroit News, gases “from the battery cells were released and ignited in the enclosed chamber. The battery itself was intact.”

…and we have not ruled out possible fraud or malicious intent. We are aware that fireworks were found in the garage in or around the vehicles. Also, an electrical panel located in the garage next to the vehicles is also being examined…

Fraud or malicious intent. That’s the phrase that Gutierrez and his attorney find offensive. Yesterday, they issued their own statement in response to Fisker’s. Some excerpts—

…Mr. Gutierrez fully accommodated the precise and somewhat peculiar demands of Fisker Automotive, who sent their self-proclaimed “SWAT Team” of engineers and inspectors (that included their own forensic cause and origin investigator) to the Gutierrez home within 24 hours of the fire.

They descended upon the Gutierrez home in alarming numbers and immediately demanded a 24-hour lock-down of his home, including the remains of the Fisker Karma vehicle. They also cordoned off portions of the Gutierrez home with non-transparent tarps to block the view from the public.

Yeah, they didn’t like the insinuation that Gutierrez may have been at fault or even behind the fire.

The family is stunned by this implication. The Gutierrez family has afforded every accommodation to Fisker and access to all evidence that public safety and law enforcement officials examined. Fisker’s statement is a grave disappointment, especially in light of the damages the family suffered and continues to suffer.

At the moment, the family is out of a home (much of it being destroyed) and three vehicles.

Okay, I’ll let you form your own opinion. But I will add that this isn’t the first time Fisker and the Karma have had troubles. There was the GM battery lab explosion, and in December a $55-million recall of faulty batteries that could lead to fire, according to Reuters.

And then the infamous breakdown of the Karma that Consumer Reports tested. (It broke down with less than 200 miles on the odometer, becoming “undriveable before it [had] finished our check-in process.”)

I’m not taking sides. Fisker is a startup bringing new, eco-friendly technology to market, which isn’t easy. The company also just lost its Department of Energy loan (for not meeting deadlines).

There’s a lot to this story. But I’ll let you decide for yourself…

Bookmark this post:

Of course right after I say all of us will be driving fully electric cars before too long (Electric Motorcycle Trailblazer), one of them goes and spontaneously combusts (allegedly), burning a whole house down.

This is a weird one, lemme just say that right now. Last week a fire burned down the Gutierrez home in Sugar Land, Texas. Since then, Jeremy Gutierrez has blamed his brand new Fisker Karma for the fire, a county fire inspector has agreed with him and Fisker has rejected both their claims.

Supposedly, Gutierrez parked his two-week old Fisker Karma (a $100,000 luxury electric sedan) in his garage. Three minutes later he smelled burning rubber. He saw the Fisker on fire and evacuated his family before the fire burned down much of the house.

After Fort Bend County head fire inspector Robert Baker examined the fire, he asserted that the Fisker was the cause. Baker said he didn’t know what exactly in the car caused the fire, but it resembled golf cart fires they have in Fort Bend County.

Here’s where it gets really weird, though. Like something out of E.T.: The Extra-Terrestrial, Fisker supposedly descended on the Gutierrez home with its own investigators, locked it down tighter than a homicide crime scene and then came out saying that the Fisker wasn’t to blame and that the company couldn’t rule out foul play (more on that later).

“I’ve worked homicide scenes with less secrecy,” Baker told AutoWeek. “There have to be about 15 engineers down here working on this one.”

The fire engulfed the garage, consuming Gutierrez’ other cars (an Acura NSX and a Mercedes SUV), and then spread to the second floor. No one was hurt.

So, Gutierrez and Baker say the Fisker started it. Fisker says it wasn’t the car and that it might be “fraud or malicious intent.” Gutierrez and his attorney don’t like what Fisker is implying. With reason, given the car’s track record so far.

Stay tuned for the rest of this post, Fisker Fire Flap…

Bookmark this post:

According to a recent survey commissioned by Nuance Research and conducted by Vocal Laboratories Inc., two out of three consumers prefer self-service over speaking to a person for customer service inquiries. The survey also showed that convenience is the largest contributing factor, with 75 percent of consumers saying that they find self-service to be more [...]

Bookmark this post:

The cloud services provider sought to add voice technologies to automate transactions.

Bookmark this post:

Can a text message a day keep patients on-track with health programs? Three years ago Doug Naegele, president of Infield Health, decided to find out.

Infield Health uses mobile technologies like SMS messaging and secure mobile web to help people adhere to programs with a certain health goal in mind. By delivering information straight to mobile phones, the company helps people better manage their health between visits to the doctor. The results, according to Infield, are better outcomes for patients and decreased costs for providers.

The company’s emphasis on mobile solutions meant finding the right telecom partner was crucial for its success. Rather than hire a developer with expertise in esoteric protocols and complex technologies, Infield chose to leverage the power of cloud communications.

Cloud communications platforms use the Internet to provide voice telephony services and text messaging capabilities to web and mobile applications. Not only does cloud communications simplify the development of telecommunications apps, it also ensures those apps can instantly scale to tens of thousands of users, or more.

Naegele recalled the first time he heard about Twilio: “I was in the office of the chief technology officer for the DC city government. There were two young programmers working on a text messaging project for parking tickets or something like that, and they were using Twilio.”

Back at home, Infield’s IT pros set up a Twilio account and started playing with the Twilio SMS API. They were immediately struck by its power and ease of use. “We can do in a day what we thought would take a month!” remarked one Infield developer.

Besides a highly reliable, scalable and secure platform, Twilio offered one big advantage to a growing company like Infield Health: It made it easy to prototype SMS applications on standard ten-digit phone numbers and move them to short codes after they were approved by a client—without any changes to the app.

Carriers require organizations that send large volumes of text messages to use short codes—five or six digit numbers that can send SMS messages at a rate of 30 texts per second. A random short code costs $3,000 for a three-month lease, and a vanity short code costs $4,500.

Naegele said it took about four months to build and test Infield Health’s first product: InfieldCMS. All Infield clients now use that content management system, built specifically for SMS and mobile web content. “Because we didn’t have to spend $50,000 integrating to a custom telecom SMS back-end with an arduous annual contract, we were able to build InfieldCMS without a first customer, then show a completed product to potential customers. As a start-up, speed-to-product was absolutely crucial,” said Naegele.

One beneficiary of the Infield/Twilio linkage is the American College of Cardiology. Infield helped ACC develop CardioSmartTXT, an app that promotes heart-healthy living. “Most of ACC’s heart-healthy outreach went from idea to testing to production-ready in under two weeks. Twilio’s flexibility makes that happen,” Naegele added.

Naegele said Twilio makes it possible for his team to “have an idea on Monday and start testing it by Wednesday.” This applies not only to features, but to entirely new apps. “When we schedule a meeting with a new prospective client, we are able to build a fully-functioning demo app and show it to them on the first day. Leading with software that works enables us to compete with firms much larger than ours,” he said.

One project that Infield Health built for WellCall, an employee wellness firm, integrates with Salesforce.com. If an employee needs a health coach, he or she simply sends a text with the word “coach” and an approved coach gets a message in Salesforce to call the employee back within one business day. “By matching a live health coach to a client who needs advice immediately, and to put it all in motion with a simple SMS, we keep our users on-track with their health goals without missing a step,” said Kerry Bradley Sylvester, director of coaching services for WellCall.

Other apps launched by Infield Health include a text message add-on to FirstBorn, which supports new mothers before and after delivery. “With Infield, FirstBorn counselors deliver healthy challenges (with incentives), patient follow-up content, and appointment reminders right to mom’s mobile phone,” said Dr. Miguel Tirado, who works with the Center for Telehealth and Cybermedicine Research at the University of New Mexico in Albuquerque, NM and the Ben Lujan Institute at New Mexico Highlands University.

Over the past year, data from Infield’s current campaigns suggest a 50 percent improvement in self-reported outcomes versus users who didn’t receive health messages. This summer, Virginia Commonwealth University will follow two Infield projects and strictly measure results.

“I love the idea of delivering health advice straight to someone’s mobile phone. Regardless of income, it’s the one device everyone has and everyone uses,” Naegele said.

Get started with Twilio SMS

Infield Health Builds Mobile Health Solutions with the Twilio SMS API

Bookmark this post:

Moms give great advice, like brushing your hair 100 times a day.

Mothers, those indispensable fountains of wisdom and life. Whether you call (a voice broadcast doesn’t count in this case) or send flowers, take a minute to remember your mom and all the things she taught you.

Ifbyphone’s VP of Marketing Jason Ferrara got a jump on the holiday when, earlier today, he shared some thoughts about motherly (marketing) advice with the readers of MarketingProfs.

What did he learn from his mom about marketing? Read on to find out:

Four Marketing Lessons From Dear Old Mom

Bookmark this post:

Voxeo Labs is looking for a DevOps/Systems Engineer to work remotely, in or around the greater Atlanta area. (We’re flexible for the right person, but an Atlanta local is highly desirable.)

Why work with us?

• We’re solving interesting problems. Tired of clients with yet another social network for cats?

• We have awesome benefits. Mind-blowing in fact. Medical, dental, the works, 100% paid for by us, even a flex card (woot!).

• We participate. Conferences, Hack-a-thons, Code Camps. You name it, we do it. Heck, even our core product is open source.

• We’re Polyglots, and love it if our team is as well. We have Java, C, C++, Ruby, Python, Groovy, node.js, C# and even ColdFusion code running in production. You’ll learn all sorts of new things.

• Hardware. Yea, you get it all (QuadCore i7 Macbook Pro, Company iPhone, etc.)

• Freedom. You can work when you want, where you want…we know you’ll get the job done!

What we want from you:

• You need to be an Ops guy/gal (duh). Send us your Github account; we really want to see your contributions to the community. Issue queues you participate in. Projects you maintain. That sort of thing.

• You need to be available for full time work. We’re planning on hiring you full time, so freelancers, development shops, and moonlighters can sit this one out.

• We are looking for an automation and change management ninja – though in the interest of full disclosure, we don’t want an actual ninja. Difficulty locating you and/or inexplicable bodies littered about the area would be difficult to manage.

• While we’re hiring someone for a Devops role, we really want more than just a one-trick pony. We’d love it if you have other skills; freelance magician, rodeo cowboy, maybe a few programming languages under your belt? We have a fair amount of Ruby code (with and without Rails), as well as some legacy JSP stuff. Got a GitHub account full of interesting code? Show us.

Want to learn more, and find out how to apply? Check out the full description on LinkedIn (and follow us on LinkedIn to keep up to date on our openings as soon as they are posted!)

©2012 Voxeo Talks. All Rights Reserved.

Related posts:

1. We’re hiring! Network Operations Cisco Engineer, Tools Engineer, NOC Engineer…
2. We’re hiring… NOC Engineer, Customer Support Engineer, Technical Account Manager, Network Security Manager…
3. We are hiring in EMEA – Business Intelligence Engineer

Bookmark this post:

Maybe the biggest knocks against all-electric vehicles at the moment are high cost, limited range, poor performance, lack of infrastructure for charging and long charging times.

All these will no doubt disappear eventually as the technology evolves—in time, everyone will own electric, I’m sure. But it takes a trailblazer to get the ball really rolling.

Oregon-based Brammo—makers of all-electric motorcycles—have created the first production electric motorcycle to go 100 miles per hour and 100 miles on a single charge. (Release the ball.)

The 100-mph mark is a meaningful one. Café racers in the 1950s—British motorcycle enthusiasts who modified production bikes to make them fast—used the 100-mph mark (known as the ton) as the ultimate achievement. If you could get your bike to do 100, you were definitely a café racer.

Times have changed since the 1950s, though (obviously). Through the second half of the last century, the café racer movement grew and (many believe) fueled the development of faster production bikes from major manufacturers like Honda and Triumph.

Today, sportbikes go a lot faster than 100 mph. There are production sportbikes that can go close to 200. So, by that standard, the Brammo Empulse is slow. But that’s not the point.

People want electric cars and motorcycles, and they want them to run like the internal combustion versions. Just like with the café racers before production sportbikes, the market is waiting for the industry to supply the product.

As a matter of fact, 100 mph is plenty fast for most of the motorcyclists out there. (Most don’t ride on the track and never go near that fast.) So, isn’t the market already there, then? Almost.

We still need to improve on the other knocks against electric. The Brammo Empulse is still like $5,000 more than regular sportbikes, isn’t as fast and takes hours to charge. But still, we’ve got a trailblazer for electric motorcycles, in any case.

Bookmark this post:

Doing pushups started as just a way for the Udemy team to stay in shape at the office. Before they knew it, the team had the crazy idea to one day turn that exercise into a way to raise money for charity. And that crazy fundraiser is about to happen.

This Thursday, San Francisco startups 99Designs, GitHub, Hotel Tonight, Udemy and Zaarly will put their hands to the ground, literally, in a pushup competition to benefit five different charities. Join us at the event May 17th, 6pm at Minna Gallery, to cheer them on! All ticket proceeds will go to the selected charities.

Five Amazing Nonprofit Organizations

Each startup chose one charity to support, which makes for an impressive lineup of philanthropic initiatives. You can support by sponsoring a team or buying a ticket to the event.

 Kiva Zip is a part of the Kiva that provides micro loans for entrepreneurs directly. Sponsor Team Github to support this nonprofit.

 Muttville is a senior dog rescue and foster service based here in San Francisco. Sponsor Team Udemy to support this nonprofit.

 DonorsChoose connects donors from around the world directly with classrooms in need of financial support. Sponsor Team 99designs to support this charity.

 CharityWater brings clean and safe drinking water to people in developing countries. Sponsor Team Hotel Tonight to support this charity.

 Livestrong, started by Lance Armstrong, leads a charge on improving the lives of those affected by cancer. Sponsor Team Zaarly to support this nonprofit.

We look forward to seeing you at the event!

San Francisco Startups Do Pushups for Charity Next Thursday

Bookmark this post:

We love when customers enhance their business in new and exciting ways with help from the Ifbyphone platform. That’s why today we’re shining light on SEO experts Pacific54. As a leading online marketing company, Pacific54 knows the importance of staying on top of a constantly evolving market. With Ifbyphone, Pacific54 is doing just that: keeping their clients current with the latest and greatest while improving the success of their business. And they’re so pleased with their success, they wanted to say a little something about it:

“We at Pacific54 developed our latest feature with the help of the Ifbyphone API. This enhancement helped us create a custom lead form that is automatically generated and sent when a visitor initiates a click-to-call request on one of our clients’ websites.

The beauty of this kind of form is while our clients are engaging with their potential customer over the phone, they receive a detailed email with all the information we could gather about that specific visitor, such as: how many times they viewed the website, what pages they viewed, their geographic location, what keyword they typed into a search to find the website, the duration of the call, and more. Think how cool it could be to answer the phone at work and already knowing what your client is looking for and how long he has been looking for it!

We’re thrilled with the results we’re seeing—we’ve found that the information from the custom form has been the key to converting more phone calls into sales—and we’re thrilled that Ifbyphone helped us achieve these results. Take a look at the form below to get a better understanding of what we’re talking about and how it can help other companies out there.”

-Avi Cohen

Avi Cohen is CEO at Pacific54, a Miami-based Online Marketing Agency focused on SEO, social media, and PPC. Avi’s expertise in Online Marketing has evolved from his passion for technology and his constant desire to implement innovative and, at times, unconventional approaches to increase his own productivity and the results he brings clients. Connect with Pacific54 on Twitter, Facebook and LinkedIn.

Bookmark this post:

The combined solution delivers enhanced regulatory compliance and quality monitoring capabilities.

Bookmark this post:

All technology, whether it is IVR (interactive voice response) or mobile applications, depends to some extent on the free flow of information.  Tech-focused companies partner, merge and share data and innovations with each other for mutual benefit.

Collaboration is key to the innovation and improvement of products and services. But what happens when this partnership sours?  This very scenario is being played out in court for a global audience, as Google and Oracle debate the origins of the Android platform, contrasting it with previously existing Java technology.

Oracle argue that Google has violated their copyright by using patented Java technology, hiring former Sun Microsystems employees and utilizing Java patents to build their Android platform.  Smartphones equipped with the Android operating system have achieved record-breaking sales worldwide. But how much of the Android platform was innovative and fresh?

Oracle argue that Google have violated their software copyrights for Java.  They filed suit against Google stating as such, and Oracle believes that they should receive damages from Google and Google should obtain a license to use Java.

As far as Google are concerned, Oracle are simply upset that they did not and have not developed technology that can rival the Android platform, and are simply trying to take down the Android platform via litigation in lieu of developing and innovating technology of their own that could viably compete for market share.

According to the New York Times, Oracle typically distribute Java free to programmers and for a nominal fee to companies, pending their agreement to abide by certain rules.  Per the article: “while other programmers and companies negotiated open source and commercial licenses, Google sought none.  Additionally, Google’s Android operating system made use of Java programming tools that were proprietary.”

Google argue that the Java programming language is freely available, and therefore they did not need to secure permissions and patents from Oracle to use the code in their Android platform.  The precedent set for the sharing and distribution of programming languages, and indeed application programming interfaces, implied consent for usage.  Open source data and coding needn’t be regulated via contract, and the proceeds a company reaps as the result of these types of mashups is exclusively theirs.

So who has the better case?  Just yesterday, a jury ruled in favor of Oracle, finding that Google did infringe on Java software code.  The damages portion of the case has yet to be litigated, so the outcome of that part of the process is what will truly determine who was correct.  If Oracle are awarded licensing fees, it would be a major victory, primarily because it could lend itself to decision-making power in future versions of Android.

This ruling could also have dramatic and long-lasting implications on how all developers use APIs, potentially making them copyrightable and thus severely curbing the free flow of information that the API promotes.  If anyone wanted to work with a company’s API in the future, they would have to navigate through a series of legal quagmires to procure patents and legal sanctions to use the code.  All information would be proprietary.

So what does this mean for the future?  Only time will tell, but all eyes will inevitably be on the outcome of this trial, primarily because of the extreme implications the outcome may have on the technology community.

Bookmark this post:

Operating in the IVR market, Plum frequently utilizes outside software to power interactive voice response applications.  Sure, we design, build and code the application, but what if you want to include text-to-speech (TTS) and automatic speech recognition (ASR) capabilities as a feature of your system?

We offer our clients the best TTS and ASR engines on the market. We partner with Nuance, AT&T natural voices and Cepstral to provide customers with the most refined and advanced technology.

In terms of system hardware components, our servers and dialogic cards included as a feature of both our hosted and onsite systems are of the highest quality, purchased from HP and Compaq, among others.

Often, we will utilize open APIs provided by various websites to combine our IVR functionality with a variety of applications and tools to create a program that includes the performance capabilities of each company.

Application programming interfaces (or APIs) enable websites and technologies to interact with one another using web service models like REST and SOAP.  Some of the most popular APIs available include the New York Times, the Guardian, Google Code, Facebook, Twitter and the Yahoo! Developer Network.

Users looking to combine their applications with these major APIs can do so easily and will find a variety of functionalities available for API creation.  Everything from article search and content storage from publications like the New York Times and the Guardian, to AdWords, Youtube, Twitter search and Yahoo messenger are features openly available for API integration.  But this availability and openness could soon be changing.

Oracle has been engaged in a protracted legal battle with Google.  The cause?  Oracle sued Google claiming that Google violated Oracle’s Java copyrights when Google built the Android operating system.

Google acquired Java patents when they purchased Sun Microsystems in January of 2010.  In addition, they hired former Sun Java engineers.  Google were also granted access to Java through a series of patents.

Oracle quickly asserted that Google’s Android platform had similarities with Oracle’s Java—based on the information above, this was no accident.  Google went out of their way to procure the information, technology and expertise of Oracle through Sun Microsystems.

Google didn’t simply study the technology that powered these devices, they mimicked it in a way that was intimately similar, so much so that Oracle felt that Google presented a pressing challenge to their business.  So they filed suit.

Stay tuned for Implied Consent?…

Bookmark this post:

…continued from Fantasyland Gets Real…

I may call Facebook Fantasyland, but a Missouri high school principal brought a heavy dose of reality to the social media website for a while. She posed as a fake student and friended her students on Facebook to…well, I don’t know…spy on them?

But the thing is, the kids went along with it. This is what I mean by Fantasyland. There’s a suspension of disbelief or something. These kids were getting friend requests from someone they obviously didn’t know and had never even heard of before, yet they were accepting.

Why were they accepting? Because they don’t refuse friend requests? Because they accepted the request out of laziness? Because they collect friends? Because maybe they’d make a new friend at school via Facebook?

I don’t know. But only the last explanation seems like a good one to me. And even then I’d have tried to find Suzy Harriston in the yearbook first.

According to Oxford University Professor Robin Dunbar, it doesn’t matter how many Facebook friends we have because our minds can only accommodate about 150 real acquaintances. Huh.

Dunbar, an evolutionary anthropologist, has studied social groupings “from Neolithic villages to modern office environments,” according to the Daily Mail. He theorizes that the “size of the part of the brain used for conscious thought and language, the neocortex, limits us to managing 150 friends, no matter how sociable we are.”

This makes sense seeing how social anthropologists often point to a 150-member size for most pre-history villages and extended families. The archeological record shows that humans have tended to live together in groups of about that size. So it makes sense that either our brains are wired that way to accommodate group size or group size accommodated our brains.

And then Dunbar looked into the Facebook thing. He wanted to see if Facebook was increasing the size of social groupings.

“The interesting thing is that you can have 1,500 friends, but when you actually look at traffic on sites, you see people maintain the same inner circle of around 150 people that we observe in the real world,” Dunbar told the Mail. “People obviously like the kudos of having hundreds of friends, but the reality is that they’re unlikely to be [more numerous] than anyone else’s.”

I guess Facebook isn’t increasing our social grouping size. Yet. But either way, I still would have checked my high school yearbook for that girl before I accepted her friend request.

Bookmark this post:

Timmy Crawford, DOer

DOer Timmy Crawford is the developer behind Cone Patrol, the Twilio-powered app serving up your powder alerts for northwest during the snow season. He is also the Lead Web Developer at The Longboard Store. Timmy recently implemented SMS reminders to customers waiting for out-of-stock items to be available and below he walks through the process on the site, as well as the conversions he’s watching.

This blog was originally posted on Timmy’s personal blog here, and you can follow him @Timmycrawford.

We have all been in that place.  After scouring the web for that latest *must have* toy, and just as you are ready to add to growing debt you see the dreaded words – OUT OF STOCK.

Retail SMS In-Stock Reminders with Twilio

Bookmark this post:

As part of Voxeo’s partnership with Integrated Voice Solutions (IVS), we’re announcing the ability to automate migration from existing legacy and end-of-life interactive voice response (IVR) systems to the more flexible and powerful open-standards technology of our communication platforms, Prophecy and VoiceObjects. When an end-of-life or legacy system is migrated to our more advanced Voxeo platform, businesses will be able to create just one single application and deploy it across any combination of voice, text or mobile web-based communication channels, giving their contact centers the ability to deliver exceptional customer experiences that aligns with customer communication preferences.

From the press release:

Businesses understand the importance of exceeding customer expectations, yet many contact centers have been stymied by the time and expense it takes to re-develop their existing voice applications to support the latest functionality and communication preferences. IVS uses Arca+ to facilitate these migrations quickly, easily and cost-effectively, which means businesses no longer have to maintain obsolete technology and pay high support and service fees.

Arca+ automates the migration from outdated, end-of-life Intervoice IQTalk and Edify EVIP platforms to Voxeo, so applications can easily be transported and maintained in a modern, standards-based development environment that mitigates the risk of future lock-in to any specific platform. The Voxeo platform supports multi-channel self-service interactions across mobile web, smartphone apps, live chat, text and social networks like Twitter. IVS will add migration options for other platforms as client needs demand. Voxeo also offers its own Migration Manager tool for Nortel/PeriPro users who want the flexibility of Voxeo’s powerful open standards technology.

“Enterprise contact centers no longer have to suffer with their existing, end-of-life IVR applications while the competition leaps ahead with cutting-edge customer outreach,” said John Amein, senior vice president of product management at Voxeo. “Arca+ provides a path for these companies to migrate to the number one platform in the communication application industry faster, easier and with less cost than rewriting.”

For more info on IVS’ Arca+ solution including demo videos of how an application can be imported into the Arca+ system, then exported to Voxeo VoiceObjects, check out our Voxeo Partner Spotlight on IVS. IVS also offers a free, no-obligation application audit and cost analysis, so companies can best understand their cost savings and options. More information on the free cost analysis can be found at http://www.voxeo.com/migration.

©2012 Voxeo Talks. All Rights Reserved.

Related posts:

1. Voxeo Partner Spotlight: IVS, Blue Flame Industries and Arca+
2. IVR For Financial Institutions: Voxeo Partners With Waterfield
3. Voxeo partners with Remego to unlock contact centers in APAC Region

Bookmark this post:

Platform includes new M*Modal Fluency Direct

Bookmark this post:

App now offers Dragon voice recognition

Bookmark this post:

Enhanced speech analytics and call recording are just some of the new features.

Bookmark this post:

Technology added to voice control

Bookmark this post:

Technology automates the transfer from end-of-life IVR systems to new platform

Bookmark this post:

Andrea Ayers to succeed Jeff Fox

Bookmark this post:

I like to call the Facebook world Fantasyland, because it’s an alternate reality where people can be whoever they want to be. I mean it as a comment on human character more than anything else.

But a principal from Missouri has showed us that Fantasyland may include actual alternate realities. And it may capture people enough that they even forget about their real lives.

According to MSNBC, Clayton High School Principal Dr. Louise Losos created a fake profile on Facebook under a false name and used it to friend students at her school. We can only guess at her motives.

In April, a Facebook member let the cat out of the bag, and the fake profile for “Suzy Harriston” went bye-bye.

“Whoever is friends with Suzy Harriston on Facebook needs to drop them. It’s the Clayton Principal,” the post declared.

The Harrison profile and its 300-plus friends vanished right after that post.

“[A] search of public records in Missouri found no results for anyone named ‘Suzy Harriston,’” MSNBC reported. Meanwhile, the school district “confirmed that no student by the name of Suzy Harriston was enrolled at the high school in the last two years.”

The School District of Clayton didn’t admit to MSNBC that Principal Losos lost her job because of the incident, but Chris Tennill, the district’s spokesperson, implied as much.

Tennill told MSNBC that Losos “has resigned as Clayton High School Principal effective June 30, 2012” and that she “had a fundamental dispute concerning the appropriate use of social media.”

The employee-student handbook for Clayton evidently forbids the use of social media between teachers and students for anything other than education-related communications.

Fair enough. I understand why the school district would ask Losos to resign, if they in fact did. What I don’t get is why the ploy ever worked in the first place.

Stay tuned for the rest of this post, 150-Friend Cap…

Bookmark this post:

A large majority of interactive voice response, or IVR, systems are equipped with automatic speech recognition technology that enables users to vocalize their input and responses, and for the machine to understand and interpret these spoken-voice prompts in turn.

Speech recognition technology acts in combination with or as an alternative to DTMF input, which collects user data input via selected numerical values from a telephone’s keypad.  The technology is novel and exciting, in that user responses and inputs are not limited to two or three preprogrammed responses, but instead enable the user to answer as they see fit (within the limitations of the ASR technology).

The one thing IVR systems often fail to take into account is the emotional range possible in responses.  In some applications this doesn’t matter (for example, applications collecting mathematical data), but for others, the ability of machines to detect specific emotions is vital to the functioning of the system and greatly affects the integrity of the data received.

An article included in the International Journal of Soft Computing and Engineering explores the process by which speech recognition engines may successfully recognize a user’s emotions.

Speech recognition applications collect user speech input and analyze features of a specific user’s speech.  Depending on what emotion they are assessing, certain features will be selected and studied to identify the emotional resonance behind spoken words. 

Per the article, a typical set of human emotions contains 300 emotional states.  To streamline the classification process, researchers identified the primary seven and split them into groups.

From this point, prosodic features associated with the seven emotions were qualified, and variables like pitch, energy and duration were taken into account for classification purposes.  Each of the 300 possible emotions fell into a range defined by the parameters of the primary seven.  From this point of feature extraction and selection, speech recognition engines can be programmed to identify emotional queues offered by the speaker.

How effective is this method in classifying and identifying emotions?  According to Ingale and Chaudhari, studies have found accuracy rates of 70% for seven emotional states, and 73% for 4 different emotional states.

This research shows that there are indeed effective techniques available for detecting and categorizing emotions within the field of speech recognition, which is guaranteed to improve as more advanced engines (like Siri) are developed and identification techniques are honed and perfecte

Bookmark this post:

As I’ve mentioned before NextUC hosts Lync. Of course after the 60 day free trial we charge for the service. Sometimes we have a need to put a customer on probation by suspending Enterprise Voice (per to peer voice and IM still work)  or suspended (noting works anymore) a customer.

So I came up with a quick and easy low impact way to suspend a Lync user so that I can lock a user out of Enterprise Voice without changing any dial plans or DIDs on the user. $UPN is the userPrincipalName (i.e. mharrison@nextuc.com) of the user.

Set-CsUser -DomainController $fqdDomainController -Identity $UPN -AudioVideoDisabled $True -EnterpriseVoiceEnabled $False

Then to re-enable the user we just do the opposite

Set-CsUser -DomainController $fqdDomainController -Identity $UPN -AudioVideoDisabled $False -EnterpriseVoiceEnabled $True

If you need to suspend a whole company (i.e. OU) so that the users can’t login to Lync you can use this command where $distinguisedName is the full LDAP path to the OU:

Get-AdUser -Server $fqdDomainController -Filter * -SearchBase $distinguishedName -SearchScope OneLevel | Disable-ADAccount

To turn a company back on you use:

Get-AdUser -Server $fqdDomainController -Filter * -SearchBase $distinguishedName -SearchScope OneLevel | Enable-ADAccount

These two commands basically get all the users (-Filter *) in an OU and pipes it to Enable-ADAccount or Disable-ADAccount.

I hope these tips help and let me know if you have questions or any tips of your own to share.

Bookmark this post:

The union will provide advanced IT and dictation solutions for the law industry.

Bookmark this post:

Companies sign interface license agreement.

Bookmark this post:

New on-demand capabilities added to identity proofing and voice biometrics platforms.

Bookmark this post:

Within IVR, speech recognition plays a pivotal roll in facilitating communication between humans and machines.  Automatic speech recognition systems are able to gather data, collect information, and process spoken user responses in an efficient and effective manner.

A highly functioning IVR application will streamline customer interactions, and make it easier and more pleasant for callers to communicate with the technology via their preferred method (which is often times speech).

When humans communicate with each other, their exchanges are laced with gestures and interpretable tones.  Body language and other mannerisms are equally important for those trying to correctly understand the true meaning and intent behind spoken words.

Machines are equipped with no such ability, but are they able to detect, understand, and interpret emotionally laced dialogue?  In an article titled Speech Emotion Recognition, researchers question whether speech recognition technology can be built with the capacity to differentiate between emotions and identify specific feelings.

Anger, happiness, sadness, surprise, and neutrality are all emotions that researchers have attempted to qualify and isolate within speech recognition systems.  The ultimate goal of this type of research is to improve human-machine communications, and make speech recognition technology more effective and efficient.

There are myriad challenges encompassed within this type of research, starting with the fact that there are few universal similarities in the way humans vocalize their emotions.  One person may sound sad by talking faintly and quietly with a downturn in their voice.  For another, this type of inflection may merely mean that they are neutral, fatigued, or contemplative.

For machines trying to establish a discernable pattern of speech recognition, this presents a variety of challenges.  There is no algorithm that can predict or process certain types of emotionally charged speech, and machines may have more difficulties interpreting the speech queues and the emotions behind them on a person-by-person basis.

As IVR applications become increasingly popular, emotion-based speech recognition becomes a highly valued feature.  Applications built to collect psychiatric information, dispense psychiatric diagnoses, operate as lie detectors, and assess mental state depend equally on the emotions behind the words being spoken, as much as they do on the actual words.

Speech-emotion recognition systems operate effectively by performing a series of tasks.  The machine receives speech input.  After this input is received, the speech recognition software performs feature extraction and selection, where speech queues are analyzed and classified for maximum interpretation effectiveness.

Feature extraction and recognition is determined mathematically, and each emotion is defined by a set of parameters designed for easy categorization.  Different emotional states are defined by variables like pitch, energy, duration, and formant

So what is the accuracy of this type of system?  Can machines really interpret how a person is feeling?

Stay tuned for A Telling Tone…

Bookmark this post:

Mayor McGinn Kicks it Off

This past weekend Startup Weekend kicked-off it’s GOV initiative in Seattle.  The premise is simple: connect communities and governments to innovative and entrepreneurial people.  Startup Weekend has done such a great job of rallying these DOers over the years in order to solve persistent and pressing problems.

Mayor McGinn and representatives from the City of Seattle, King County and the State of Washington were on hand this past Friday to kick-off the event and thank everyone for spending their weekend thinking about how to use technology and think differently about improving their communities.

Crunching Data and Hacking Government

Team My Spot busy finding a better way to pay for parking

Over 100 developers, designers and entrepreneurs participated this event, which was held in Seattle’s City Hall.  This was my first ever Startup Weekend and I was excited to be a mentor and help the teams define their vision, compile customer validation and build their prototypes.  There were several other mentors there from both the public and private sectors.  A big focus of the weekend was data: the government has a ton of it, but it’s not necessarily being put to productive use.  Teams were challenged to take advantage of data that was currently being made available by companies such as Socrata. For governments who haven’t yet released data sets, teams tweeted the hashtag #needdatanw to request information.

Which Bus Makes Finding the Right Bus Safer and Easier

Gilad Gray and team tied for 1st with Which Bus

By the end of the weekend dozens of ideas had been whittled down to 10 teams that would present to the judging panel, which consisted of Bill Schrier (City of Seattle CIO), Kate Matsudaira (Decide.com), Bharat Shyam (State of Washington CIO), Greg Gottesman (Madrona) and Mike Mathieu (Front Seat).

I was fortunate to spend some time working with a team building Which Bus, an app that takes the best of Google Maps and One Bus Away providing a simple way to figure out how to get from point A to B in Seattle. Figure out how to use public transit, see which bus to hop on and when it’s expected to arrive.  A big theme of the weekend was accessibility and making sure that public services are available to as many people as possible.  Most apps and services require a web browser or smart phone, which excludes many of the residents of the State of Washington.  Once Team Which Bus decided to use Twilio for SMS notifications, it only took about an hour to implement, which caught the attention of Sol Villarreal, the community engagement director for the City of Seattle.

Thinking about technical solutions for government issues was new to me and my guess is it was new to many of the attendees of Startup Weekend GOV.  I’m really happy that I had the chance to attend and am looking forward to see this initiative roll-out to more cities and hopefully see some great solutions get some tractions and improve the quality of our communities.

Startup Weekend Launches its GOV Initiative in Seattle

Bookmark this post:

Interactive voice response (IVR) applications are dependent on automatic speech recognition technology (ASR) to function at the highest levels.  IVR systems are typically developed under the assumption that the technology will be customer facing and interact with numerous individuals in a variety of different situations and scenarios.

IVR applications are often times built with speech recognition technology included as one of the main features of the system.  Instead of exclusively relying on DTMF recognition (which is defined as the system’s ability to process numerical entries from a keypad) applications equipped with ASR functionality allows respondents to speak their answer.

The front-end technology involved in this type of system is very easy to understand, and when it functions correctly, its usage is mostly self-explanatory.  Delving a bit deeper however, one immediately realizes how complicated and challenging speech recognition truly is.

The most effective speech recognition systems are equipped with fairly expansive vocabularies, equipping the system with the ability to interpret a variety of the words that might be spoken in response to any given prompt.

While it is nearly impossible to quantify the actual number of words in any give language, a 2010 study conducted by Harvard and Google found that the English language contains around 1.2 million words, with the lexicon growing by 8,500 words per year.

Speech recognition technology is constantly and continuously improving, but even with huge developmental strides made in the software that determines vocabulary capacity, it would be impossible to include even half of the potential words that may be uttered by a speaker.

In addition, there are a variety of pronunciations, accents, and inflections included in human speech that speech recognition technology may be unable to account for.  Language caveats like homonyms, confusable, and similar sounding words add an additional layer of difficulty.

Alliteration, idioms, and other grammatical anomalies may not present a challenge when they occur in everyday speech, but can make language interpretation very difficult for software applications.

So how do you get a machine to not only correctly interpret what you are saying, but to also detect the context in which you are saying it?   Is it even possible?  The International Journal of Soft Computing and Engineering recently published an article about emotion perception within speech recognition applications, which explored this very subject.

Stay tuned for You Are Either Angry or Happy…

Bookmark this post: